public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Petr Mladek <pmladek@suse.com>
To: Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@kernel.org>,
	Peter Zijlstra <peterz@infradead.org>
Cc: Laurence Oberman <loberman@redhat.com>,
	Vincent Whitchurch <vincent.whitchurch@axis.com>,
	Michal Hocko <mhocko@suse.com>,
	linux-kernel@vger.kernel.org, Petr Mladek <pmladek@suse.com>
Subject: [PATCH 2/3] watchdog/softlockup: Report the same softlockup regularly
Date: Mon, 19 Aug 2019 12:47:31 +0200	[thread overview]
Message-ID: <20190819104732.20966-3-pmladek@suse.com> (raw)
In-Reply-To: <20190819104732.20966-1-pmladek@suse.com>

Softlockup report means that there is no progress on the given CPU. It
might be a "short" affair where the system gets recovered. But often
the system stops being responsive and need to get rebooted.

The softlockup might be root of the problems or just a symptom. It might
be a deadlock, livelock, or often repeated state.

Regular reports help to distinguish different situations. Fortunately,
the watchdog is finally able to show correct information how long
softlockup_fn() was not scheduled.

Report before this patch:

[  320.248948] watchdog: BUG: soft lockup - CPU#2 stuck for 26s! [cat:4916]

And after this patch:

[  480.372418] watchdog: BUG: soft lockup - CPU#2 stuck for 26s! [cat:4943]
[  508.372359] watchdog: BUG: soft lockup - CPU#2 stuck for 52s! [cat:4943]
[  548.372359] watchdog: BUG: soft lockup - CPU#2 stuck for 89s! [cat:4943]
[  576.372351] watchdog: BUG: soft lockup - CPU#2 stuck for 115s! [cat:4943]

Note that the horrible code never really worked before the accounting
was fixed. The last working timestamp was regularly lost by the many
touch*watchdog() calls.

Also note that the full report is useful to distinguish livelock
and deadlock.

Signed-off-by: Petr Mladek <pmladek@suse.com>
---
 kernel/watchdog.c | 25 +------------------------
 1 file changed, 1 insertion(+), 24 deletions(-)

diff --git a/kernel/watchdog.c b/kernel/watchdog.c
index bd249676ee3d..2058229ed398 100644
--- a/kernel/watchdog.c
+++ b/kernel/watchdog.c
@@ -173,10 +173,8 @@ static DEFINE_PER_CPU(unsigned long, watchdog_period_ts);
 static DEFINE_PER_CPU(struct hrtimer, watchdog_hrtimer);
 static DEFINE_PER_CPU(bool, watchdog_restart_period);
 static DEFINE_PER_CPU(bool, softlockup_touch_sync);
-static DEFINE_PER_CPU(bool, soft_watchdog_warn);
 static DEFINE_PER_CPU(unsigned long, hrtimer_interrupts);
 static DEFINE_PER_CPU(unsigned long, soft_lockup_hrtimer_cnt);
-static DEFINE_PER_CPU(struct task_struct *, softlockup_task_ptr_saved);
 static DEFINE_PER_CPU(unsigned long, hrtimer_interrupts_saved);
 static unsigned long soft_lockup_nmi_warn;
 
@@ -268,7 +266,6 @@ static void __touch_watchdog(void)
 {
 	__this_cpu_write(watchdog_touch_ts, get_timestamp());
 	__restart_watchdog_period();
-	__this_cpu_write(soft_watchdog_warn, false);
 }
 
 /**
@@ -429,31 +426,13 @@ static enum hrtimer_restart watchdog_timer_fn(struct hrtimer *hrtimer)
 		if (kvm_check_and_clear_guest_paused())
 			return HRTIMER_RESTART;
 
-		/* only warn once */
-		if (__this_cpu_read(soft_watchdog_warn) == true) {
-			/*
-			 * When multiple processes are causing softlockups the
-			 * softlockup detector only warns on the first one
-			 * because the code relies on a full quiet cycle to
-			 * re-arm.  The second process prevents the quiet cycle
-			 * and never gets reported.  Use task pointers to detect
-			 * this.
-			 */
-			if (__this_cpu_read(softlockup_task_ptr_saved) !=
-			    current) {
-				__this_cpu_write(soft_watchdog_warn, false);
-				__restart_watchdog_period();
-			}
-			return HRTIMER_RESTART;
-		}
-
 		if (softlockup_all_cpu_backtrace) {
 			/* Prevent multiple soft-lockup reports if one cpu is already
 			 * engaged in dumping cpu back traces
 			 */
 			if (test_and_set_bit(0, &soft_lockup_nmi_warn)) {
 				/* Someone else will report us. Let's give up */
-				__this_cpu_write(soft_watchdog_warn, true);
+				__restart_watchdog_period();
 				return HRTIMER_RESTART;
 			}
 		}
@@ -461,7 +440,6 @@ static enum hrtimer_restart watchdog_timer_fn(struct hrtimer *hrtimer)
 		pr_emerg("BUG: soft lockup - CPU#%d stuck for %us! [%s:%d]\n",
 			smp_processor_id(), duration,
 			current->comm, task_pid_nr(current));
-		__this_cpu_write(softlockup_task_ptr_saved, current);
 		print_modules();
 		print_irqtrace_events(current);
 		if (regs)
@@ -483,7 +461,6 @@ static enum hrtimer_restart watchdog_timer_fn(struct hrtimer *hrtimer)
 		add_taint(TAINT_SOFTLOCKUP, LOCKDEP_STILL_OK);
 		if (softlockup_panic)
 			panic("softlockup: hung tasks");
-		__this_cpu_write(soft_watchdog_warn, true);
 	}
 
 	return HRTIMER_RESTART;
-- 
2.16.4


  parent reply	other threads:[~2019-08-19 10:47 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-08-19 10:47 [PATCH 0/3] watchdog/softlockup: Make softlockup reports more reliable and useful Petr Mladek
2019-08-19 10:47 ` [PATCH 1/3] watchdog/softlockup: Preserve original timestamp when touching watchdog externally Petr Mladek
2019-10-21 12:42   ` Peter Zijlstra
2019-10-21 13:04     ` Petr Mladek
2019-08-19 10:47 ` Petr Mladek [this message]
2019-10-21 12:43   ` [PATCH 2/3] watchdog/softlockup: Report the same softlockup regularly Peter Zijlstra
2019-10-21 13:40     ` Petr Mladek
2019-10-21 14:09       ` Peter Zijlstra
2019-08-19 10:47 ` [PATCH 3/3] Test softlockup Petr Mladek
2019-10-21 12:45   ` Peter Zijlstra
2019-10-21 13:06     ` Petr Mladek
2019-10-21 12:32 ` [PATCH 0/3] watchdog/softlockup: Make softlockup reports more reliable and useful Petr Mladek
  -- strict thread matches above, loose matches on Subject: below --
2019-06-05 14:09 [RFC " Petr Mladek
2019-06-05 14:09 ` [PATCH 2/3] watchdog/softlockup: Report the same softlockup regularly Petr Mladek

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20190819104732.20966-3-pmladek@suse.com \
    --to=pmladek@suse.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=loberman@redhat.com \
    --cc=mhocko@suse.com \
    --cc=mingo@kernel.org \
    --cc=peterz@infradead.org \
    --cc=tglx@linutronix.de \
    --cc=vincent.whitchurch@axis.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox